Notes on 'Good' Research Questions
- The majority of research projects follow a similar structure. Imagine the structure of a research project taking an hourglass form:
Broad questions in area of interest [How can we encourage economic growth?] -> Narrowed/Focused Questions [Ethically, can more weight be given to economic growth over the other principles of sustainable development] -> Observation -> Analysis -> Research Conclusions -> Generalise back to Questions [Ethically can we encourage economic growth without compromising long term goals?]
- A large number of social research projects use a general problem as a starting point which is usually broad and cannot be tackled in a single research study [The current economic issues with the UK desperate for economic growth]. This leads to a narrowed question in the context of existing theory in order to address the problem [Ethically, can more weight be given to economic growth over the other principles of sustainable development] Often this question is still broad and requires a hypothesis to describe in detail what will happen in the study.
- In causal studies the variables of cause and effect are important along with the distinction between the two. There are two types of causes, ones the researcher can control and the ones they can't, whereas the effect is the outcome.
- Research in social sciences is always conducted in the social context where people are observed, measured and questioned. Unfortunately in most cases not everyone can be involved in a research project there for we must sample a representative group of people. The distinction between the theoretical population and the final sample needs to be made.
-In causal studies the focus is on the effects of a cause on an outcome which are directly related to the research question. A study like this would contain the following components:
-> The research problem
-> The research question
-> The cause
-> The people
-> The effect
-> The design
Summary of Grix (2002) (need to include specific references to Grix paper)
The article Introducing Students to the Generic Terminology of Social Research by Jonathan Grix has two significant aims; the first assisting students and academics to understand basic terminology used within social science research and the second enabling a clear understanding of the interrelationships between the core concepts of research in social sciences. Academics often use alike terms but associate them with unalike meanings leading to confusion, within both students and academics. In order for constructive discussion to take place students must understand not only the terminology of core concepts but also their relationship to one another. There are five core concepts of social science research; ontology is the assumption of what exists; epistemology is the discussion of what we can know about what exists; methodology refers to the limitations and potentialities of methods; methods is concerned about the techniques used to collect and analyse data and sources is the consideration of where data should be collected from. The importance of this understanding is demonstrated by the ESRC (Economic and Social Research Council) urging for more comprehensive postgraduate training in research. Grix compared the terminology of social research to that of a bricklayer where a wall cannot be built without the full understanding of basic tools, how to use them and in what order. Any social science research needs to be supported by a transparent understanding of ontological and epistemological perspectives to defend our positions, recognise those of others, provide clarity discussing theory and social phenomena and to understand the crucial relationships between different components of research. In most cases ontology will be the starting point of research based on assumptions made social reality. An ontological position will be significantly affected by the researcher's cultural and academic context which can be see within objectivist and constructivist perspectives. Objectivism sees social phenomena independent of social actors whereas constructivism views social actors continually carrying out social phenomena. In order to consider an epistemological perspective the nature of social and political reality needs to be investigated. Epistemology is concerned with the theory of knowledge therefore is an important branch of philosophy with the intention of constructing new theories over those that already exist adapting to different contexts of research. For instance Grix questions if it is possible to apply a theory, developed within a western culture, to explain social phenomena in an Eastern European transition state. An epistemological position of a positivist (application of methods to grasp social reality) will require a contrasting methodology compared to an interpretivist perspective (utilises the subjective understanding of a social scientist) leading to different observations of social phenomena. Grix uses Plato's cave analogy to differential ontology and methodology which illustrates how people think directed by social and cultural norms, how different perspectives exist and alternative ways of knowledge gathering. It is vital that students understand how the researcher's experience and perspective impacts upon the entire research process. Grix provides a diagram of how the core concepts are linked disagreeing with method led research. I agree with Grix, as he pointed out this can easily lead to the incompatibility of the research methods and questions weakening the interconnectivity of core research concepts. Some research methods are favoured over others but this will not necessarily lead to good research. This can be achieved through applying all concepts appropriately and in order. These concepts are interconnected but this does not imply that one determines the other as Grix demonstrates using Putnam School's approach to the social capital debate taking a foundationalist ontological perspective and a positivist epistemological position but if you alter either perspective then this will lead to different research methods and sources. Grix concludes that this article is not encourage people to agree with his arguments but to demonstrate the impact of different research origins which in turn lead to different research strategies.
Data task 1 - see separate file
Data Task 1 -see separate file
Review of Nobel et al. (2004)
The aim of the paper Measuring multiple deprivation at the small-area-level, written in 2004 by Michael Noble, Gemma Wright, George Smith and Chris Dibben, is to collate the key aspects which support the indices of multiple deprivation; focusing on innovative methods employed in its modelling and the statistical processes behind the production of indices in order to measure multiple deprivation at the small-area level.
A background of area-level multiple deprivation is discussed indicating that the indices acknowledged in this paper are compositional in definition. This is an area identified as being deprived due to the high proportion of deprived people inhabiting the area. This paper adheres to a particular conceptual framework underpinned by deprivation being suffered by the individual as it is theoretically possible to collate measurements at an individual level to consider the entirety of deprivation. This argument is presented strongly referring to relevant literature highlighting that areas identified as being deprived could contain a majority of individuals experiencing low levels of deprivation and vice versa. This is because the majority of people who experience deprivation do not tend to live in areas identified as deprived.
The theoretical framework that underpins the measure of small-area deprivation presented here is utilised to give each indicator a weighting so indices can interact and the effect of two or more forms of deprivation can be measured. This can cause ethical problems as it can often be difficult to decide which aspects of deprivation should be assigned more weight than the others. There are five approaches to this; driven empirically, by literature, relevance to policy, arbitrary decisions or by consensus. Weighting causes the most significant issues when tackling forms of social deprivation as these are often subjective and difficult to measure. It is argued logically that even if weighting was dismissed this then implies that all forms of deprivation are equal thus someone with low income and poor health is equally deprived as an individual experiencing housing stress with poor access to services.
The methodology utilised in this paper consists of the identification of deprivation measures that can be reasonably combined demonstrated by measuring them accurately and testing for fitness. If any error arises this must be dealt with avoiding unintentional effects. The measure of each domain needs to have the capacity to be ranked and then standardised in order to form part of an index of deprivation with explicit weights.
The scale of sampling here is small, looking individually constructing a measure for small-area multiple deprivation therefore the reliability of a measure can easily be questioned. This has led to the use of shrinkage to make the resultant measures more robust. This involves adjusting unreliable scores towards a most robust score, for example the district or national mean.
Throughout the paper the authors discovered that measuring deprivation at the small-area level requires a transparent conceptual model which produces, using the best accessible data, a series of weighted measures. It was identified that the measures of small-area deprivation must be driven by the chosen conceptual framework not by data or statistical techniques. Weaknesses in the methodology were also highlighted needing further research refining the shrinkage estimation technique, exploring the areas from which strength should be borrowed to support robustness of indicators and selecting the appropriate domain weights.
Critical Review of Survey 1
a) This survey has largely utilised open ended questions as the information gathered will be used to identify any design problems which have not been predetermined. Open ended questions pose a problem for a backpacker who is highly mobile and would be best suited to closed questions which are quick to answer and easy to understand. Although this method may only highlight obvious design problems it should improve the response rate. The postcards were distributed disproportionately so data collected from different locations will be harder to compare. For example comparing Brisbane, where 2 hostels distributed the survey and Melbourne, where only one hostel distributed the survey but location could be irrelevant in this instance as the backpackers were asked not to worry about filling in the postcard in any particular location. The hostel staff were told to hand a postcard to each incoming guest and the hostel only had a limited supply of postcards therefore it would increase response rate if each guest was only given a card if they were willing to fill it in. There was a significantly disproportionate amount of surveys handed out by travel agencies compared with hostels, this could have an impact upon responses as a different type of backpacker may use a travel agency. If travel agencies were to distribute surveys then they should distribute a share which is equal to that of hostels. There were three versions of the survey circulated, this can cause issues when collating and comparing the feedback data especially when only 800 postcards were distributed in different locations making.
